Ground Probe
Ground should fall between 0 and 180 degrees. In air, when you zero the detector, you will see phases of both positive and negative values - those are just noise and don't pay attention to them. If you are getting a negative phase for any of the frequencies when you lower the loop to the ground however, you probably have an overload problem. Most ground that we see is 178 (-94) to 165 degrees except salt which falls close to 90 degrees (VDI 0).
Nonferrous targets (coins and such).VDI 0 - +95
The VDI reading is converted from the ground phase normalized to standard VDI units. It's either going to be from the strongest signal if the detector is running best data, or the average of the two strongest signals if correlate is selected. If you have ground coming in at -94(VDI) with a signal strength of 2% and another ground coming in at -91 with a signal strength of 7%, it would seem to me that the ground coming in at -91 is more mineralized even though it is a less negative VDI. 0-10%=low mineralization, 11-25%=moderate and 25% on up = high/very high. The lower the mineralization, the slower one should sweep - which may mean a lower filter should be selected. The phase tells you composition (between purely ferrous and purely salt), while the signal strength tells you the concentration (more mineralized).
Sensitivity Zoom
In the Sensitivity Zoom window, you see the signal strength at a percentage of when the detector overloads. This isn't on any specific frequency - the reading is on the incoming signal before it has been separated into the 3 frequencies. The goal is to keep the ground signal about 10% so that you still get depth and can still see shallow targets without overload. Signal% doesn't indicate signal "loss", but rather the amount of residual signal present. If ground + null are the cause, then this doesn't take away from sensitivity. Instead, it takes away from large/shallow target response, manifested in a signal overload. None of these are losses, but rather they reduce the available dynamic range. The lower this % number the stronger the signal will come through or be received.
The Noise% measures the External Interference. 0% or a low number indicates very little external noise and low ground interference, thus allowing for using a higher preamp RX Gain. A high % reading of Noise of 50% or more would require you to:
Reduce the RX Gain
Use a different frequency on your program
Use a smaller search coil.
Cranking up the Gain or Sensitivity can obliterate a good signal with an increase in noise. Particularly small and deep targets get the worse treatment. EMI shows up as chatter and reduces the low end, the minimum detectable signals. Ground causes overload or makes GB difficult and the reduction is on the high end, large/shallow target overload.
The V3 (not V3i) recommended a preamp gain to keep the total residual signal level below 10%. However, 10% is a rather arbitrary level. You can run your V3 at several notches above the recommended preamp gain. I would go no more than 30-40%. Quite often EMI will be the limiting factor. If EMI is the limiting factor, run the preamp gain as high as you can, but keep it stable. EMI-induced falsing can also be reduced by lowering the DISC sensitivity, and often this is a better way to reduce EMI falsing than lowering the preamp gain. Also, when EMI limits the max gain TX boost becomes an effective way to get some extra depth. The rule of thumb is that, to gain another inch of depth the gain must be doubled. This means that increasing the gain from 14 to 15 will not improve depth much at all whereas increasing the gain from 2 to 4 should gain another inch. So 4 to 8 gains another inch and 8 to 16, oops no 16, another inch. So don't get crazy on setting the RX at the upper levels because it provides diminishing rewards. In some cases the added noise will actually negate the increase. So in my opinion, higher RX and more noise and signal loss would be about the same as lower RX with less signal loss. If you are hunting in a clean area with only deep targets, then run the gain just below the overload point for max depth. But if there are a lot of shallow targets, then this will result in a lot of audio overload signals which might be annoying. The bottom line is, with any coil, run the gain as high as you can for stable operation, just as you'd do with any detector.